Genomic heterogeneity within B/T mixed phenotype acute leukemia in a context of an immunophenotype

Highlights • Molecular heterogeneity within B/T mixed phenotype acute leukemia (MPAL) in the context of an immunophenotype.• T-lineage predominant B/T MPAL sharing the genetic features of T-ALL.• Patients generally respond well to lineage matched treatment.• Future studies to explore phenotype-genotype associations and its implication on diagnosis and treatment choice are needed.


Introduction
Mixed phenotype acute leukemia (MPAL) is a heterogenous group of aggressive acute leukemia (AL) where the blast population(s) express more than one type of lineage-defining marker and the leukemia cannot be classified into other clinically or genetically defined categories such as acute myeloid leukemia (AML) with myelodysplasia-related changes (AML-MR) by the WHO classification, including the recently published 5th WHO classification and the International Consensus Classification (ICC) [1][2][3][4][5].AML-MR includes cases with myelodysplasia-related cytogenetic abnormalities (including a complex karyotype) and gene mutations.Diagnosis and clinical management of MPAL are often challenging due to rarity of the disease (accounting for ~3% of AL), complexity of the lineage marker expression with associated evolving diagnostic criteria, and genomic diversity.
While the aforementioned studies suggest that genetic aberrations of B/T MPAL are similar to those of T-lymphoblastic leukemia/lymphoma (T-ALL), given its rarity with only a few reported studies, we sought to investigate the clinicopathologic and genomic characteristics of 3 cases of B/T MPAL at our institution.We were particularly interested in exploring whether the genomic landscape is associated with an immunophenotype of T-lineage or B-lineage predominance given the distinctly different pathogenesis between T-ALL and B-lymphoblastic leukemia/lymphoma (B-ALL).Evaluation of these genotype-phenotype (lineage predominance) associations in B/T MPAL has not yet been systematically performed.Our study is the first to apply the 5th WHO and ICC classifications in diagnosis and to demonstrate molecular heterogeneity within B/T MPAL in a context of T-lineage or B-lineage predominance.Specifically, T-lineage predominant B/T MAPL shares a genetic signature of T-ALL whereas B/T lineage co-dominant B/T MPAL lacks such a T-ALL signature.Furthermore, we report a previously underreported SFPQ::ZFP36L2 fusion in B/T MPAL and a novel NUP98:: MLLT1 fusion, thus providing further insight into MPAL pathogenesis.Lastly, we discuss the challenges in applying the MPAL diagnostic criteria as to whether the cases with immunophenotypically defined MPAL and carrying myelodysplasia-related genetic changes should be diagnosed as MPAL or AML-MR.

Clinical and laboratory data
We identified 3 cases of de novo newly diagnosed B/T MPAL with comprehensive flow cytometric immunophenotypes, cytogenetic studies, and next generation sequencing (NGS) between 2019 and 2022.The clinical and laboratory data of the patients (Table 1) were retrieved from the electronic medical record system in accordance with an IRBapproved protocol.

Morphology and flow cytometry
Wright-Giemsa-stained peripheral blood and bone marrow aspirate smears and hematoxylin/eosin-stained trephine biopsy/aspirate clot sections and lymph nodes were reviewed.Flow cytometric analyses were performed on bone marrow aspirate, peripheral blood, or lymphoid tissue on the 10-color BD FACS Canto instruments (Becton Dickinson) and analyzed using cluster analysis with Cytopaint Classic software (Leukobyte) as described preciously [12].A panel of lymphoid and myelomonocytic markers (Table 2) was used for comprehensive immunophenotypic analysis.A total of 100,000 events/tube was collected.
B/T MPAL was diagnosed per the 5th WHO/ICC classifications for hematolymphoid tumors [3].Specifically, strong antigen expressions of B-lineage defining markers (CD19, CD22, and CD10) and T-lineage defining markers (surface CD3 and intracellular CD3) were defined as the brightest antigen expression in a population of blasts exceeding an intensity level of 50 % of normal mature B cells and T cells, respectively.In this study, the brightest antigen expression in all cases reached the intensity of their normal counterparts.Additionally, all cases exhibited a pattern of heterogenous antigen expression.

Chromosomal and FISH analyses
Karyotypic analysis was performed on Giemsa-banded metaphase cells from bone marrow aspirates using standard techniques.Fluorescence in situ hybridization (FISH) was performed on either bone marrow aspirates or lymphoid tissue using standard techniques.FISH panel details are presented in the footnote of Table 3.

Next generation sequencing
DNA and RNA were isolated from either fresh bone marrow (cases 1 and 3) or lymph node (case 2) with paired germline specimens from saliva.A custom institutional next generation sequencing (NGS) panel with 138 cancer-related genes in DNA and 1526 genes in RNA was performed on Illumina NextSeq 550 instruments.The variant allele frequency limit of detection was 5 % for single nucleotide variants and 10 % for indels.RNA sequencing required 10 split or discordant reads for gene fusions.Full details of the genes tested, exon coverage, and the bioinformatics pipeline are available at http://www.utsouthwestern.edu/sites/genomics-molecular-pathology/.
The patient was treated with a T-ALL-based protocol and achieved  1).

Case 2
A 41-year male presented with an enlarged lymph node in the neck.Imaging studies revealed lymphadenopathy above and below the diaphragm (stage IV).Lymph node biopsy showed a lymphoblast infiltrate (Fig. 2).Flow cytometric analysis on lymph node revealed a T/B lymphoblast population with a T-lineage predominance, positive for cCD3, CD19/CD22 (subset strong), but negative for CD1a, sCD3, MPO or monocytic markers (Table 2).This immunophenotype fulfilled the diagnostic criteria for B/T MPAL (biphenotypic) with a T-lineage predominance and ETP-ALL-like immunophenotype.BM marrow was involved by B/T MPAL with a low percentage of lymphoblasts (5.3%).Cytogenetic study revealed a normal male karyotype.NGS identified multiple gene mutations in JAK3, STAT5B, PHF6, NRAS, and RUNX1 (Table 3).
The patient was initiated on T-ALL therapy following the CALGB 10403 protocol and achieved partial remission with lymph node involvement at the end of induction (despite MRD negative in BM) and achieved complete remission with no lymph node involvement at the end of the second course of chemotherapy.He underwent allogeneic hematopoietic stem cell transplantation (allo-HSCT) and remained in complete remission at the last follow-up (12 months after diagnosis) (Table 1).
The patient was treated with B-ALL-based protocol and achieved complete remission at the end of induction without minimal residual disease (MRD < 0.01 % by flow cytometry on BM sample) and was in first complete remission at the last follow-up (27 month after diagnosis).
ETP-ALL-like IP: early T-cell precursor ALL (ETP-ALL)-like immunophenotype; c, cytoplasmic; s, surface.All cases demonstrated a pattern of heterogenous antigen expression (e.g., coordinated strong co-expression of B-cell markers (CD19 and CD22) on a subset of blasts in case 2.

Discussion
This is the first study demonstrating molecular heterogeneity within B/T MPAL in a context of an immunophenotype of T-lineage or B-lineage predominance using the 5th WHO and ICC diagnostic criteria.Specifically, T-lineage predominant B/T MPAL shares a genetic signature with T-ALL, whereas B/T lineage co-dominant B/T MPAL lacks such a T-ALL signature.Importantly, all three patients responded well to lineagematched ALL-based therapy.
Such T-ALL-like genetic alterations include mutations in PHF6, There are previously underreported and novel genetic alterations in our cohort that merit further discussion.In the two cases of T-lineage predominant B/T MPAL, we identified alterations in STAT5B and SFPQ:: ZFP36L2 which have been previously reported in a small subset of T-ALL

Table 3
Cytogenetic and molecular findings of B/T MPAL.
Importantly, while our finding of a phenotype-genotype association within B/T MPAL in a small cohort requires further investigation of genomic aberrations in the context of lineage predominance, such association may have implications on diagnosis and treatment.Current investigative studies on the genomic landscape of MPALs mainly focus on genetically defined entities, such as KMT2A-r, BCR::ABL1, the recently added ZNF384-r and BCL11B-activated MPAL, and generic MPAL-NOS (not otherwise specified, by immunophenotype) without providing immunophenotypic specifics on which lineage predominates [7][8][9][10][11]21,25,26].Likewise, MPAL diagnosis may not include such immunophenotypic detail.We strongly advocate for including immunophenotypic-lineage predominant information in clinical MPAL diagnosis and in research studies for two main reasons.First, molecularly defined MPAL may have heterogeneous genetic features, e.g., BCR:: ABL1 MPAL encompassing two genetic subgroups (i.e., a G5 subtype with an AML-like gene expression profile and a G8 subtype with common lymphoid progenitor disease-like signatures) [9].Second, lineage assignment is a fundamental attribute in tumor classification [3] and many tumor regimens are tailored to lineage assignment while genomic/molecular attributes are important for tumor biology and therapeutic targets.It has been shown that MPAL patients who received lineage-matched therapy have a better clinical remission rate compared to those who did not [21].The importance of lineage matched therapy is supported by the fact that all three patients in our cohort with B/T MPAL achieved complete remission after receiving lineage-matched ALL-based therapy.Following current treatment guidelines [25], one adult patient in first complete remission (CR1) underwent allo-HSCT while allo-HSCT for the two pediatric patients in CR1 was not indicated.These data advocate that clinical diagnosis and investigative studies of MPAL should include phenotypic description along with lineage predominance as well as genetic information.
Lastly, myelodysplasia-related cytogenetic and gene mutations in cases of immunophenotypically defined B/T MPAL raise diagnostic challenges.Complex karyotype, myelodysplasia-associated chromosomal abnormalities (-5q, +8), and myelodysplasia-associated gene mutations (as defined by 5th WHO and ICC, such as mutations in ASXL1, BCOR, RUNX1, SF3B1, U2AF1, and EZH2) have been reported in ~30%, ~10-20%, ~30% of B/T MPAL cases, respectively [7][8][9][10][11].The 3 cases of B/T MPAL in our cohort harbored such genetic/molecular profiles (complex karyotype, mutations in BCOR and RUNX1, one in each case).These scenarios create diagnostic controversy as to whether these cases with immunophenotypically defined B/T MPAL (but without evidence of strong myeloid differentiation) and carrying myelodysplasia-related genetic changes should be diagnosed as MPAL or AML-MR.This diagnostic challenge was also discussed in the two recent reviews [27,28].Under the current diagnostic guidelines by the 5th WHO and the ICC [3][4][5], such genetic profiles necessitate a diagnosis of AML-MR rather than B/T MPAL.Notably, emerging evidence would argue for a diagnosis of B/T MPAL for multiple reasons.First, similar to our cohort, no lineage defining myeloid markers were expressed in any of these reported cases of B/T MPAL harboring myelodysplasia related genetic abnormalities.Second, such B/T MPAL patients generally respond well to ALL-based regimens.Third, myelodysplasia-associated gene mutations are not specific for AML-MR and present in ~30% of de novo AML cases without myelodysplasia-related changes in a recent study [29].Furthermore, these mutations are not even specific for AML, e.g., EZH2, ASXL1, and RUNX1 mutations are present in ~5% of T-ALL and in ~2% of B-ALL [30,31].Fourth, nearly all recent genomic/clinical studies that have defined MPAL in the last 5 years since publication of the 2017 WHO classification include those cases with so-called myelodysplasia defining cytogenetic/molecular features [7][8][9][10][11]21,25,26]; this apparent reluctance to fully apply the WHO criteria is a testament to this unsettled debate.Future well-designed studies to include comprehensive phenotypic features with assessment of lineage predominance as well as genetic features are imperative to reconcile diagnostic controversy and refine MPAL diagnostic criteria.
In summary, our study demonstrates molecular heterogeneity within B/T MPAL in a context of an immunophenotype of T-lineage or B-lineage predominance in a cohort of 3 cases.Future studies are needed to evaluate phenotypic and genotypic associations in MPAL including assessing lineage-matched treatment in the context of genetics.A better understanding of MPAL biology is needed to clarify diagnostic ambiguity and to refine diagnostic criteria centered on myelodysplasiaassociated cytogenetics and gene mutations.

Funding
None.

Ethics statement
This study was approved by the institutional review board.Ethics approval and/or informed consent were not required for this study.

Informed consent
This study was approved by Institutional Review Board.Informed consent was not required per the study protocol.

NOTCH1,
and in genes involving the JAK-STAT (JAK3 and IL7R) and the RAS (NRAS, PTPN11, and NF1) pathways, as well as fusion of SET:: NUP214.Notably, this genetic profile was present in the two cases of B/ T MPAL with a T-lineage predominance in our study.Intriguingly, our study also highlights a previously underreported finding of B/T-lineage co-dominant MPAL harboring genetic aberrations distinct from those of T-ALL; this case (patient 3) harbored mutations in NF1 and BCOR1 along with NUP98::MLLT1.Our results suggest molecular heterogeneity within B/T MPAL in a context of an immunophenotype of T-lineage versus B-lineage predominance, which is not entirely unexpected given the distinctly different pathogenesis of T-ALL versus B-ALL[3,14].Indeed, one reported case of B/T MPAL (1/9) without specifics on the T-lineage or B-lineage predominance had IKZF1 mutation[10], a common alteration in Philadelphia chromosome-like (Ph-like) B-ALL[15].
but have not yet been reported in B/T MPAL.Concurrent mutations in RUNX1 and PHF6 along with NOTCH1 mutation indicate an alteration of early hematopoietic differentiation to promote T-cell differentiation at the expense of B-cell differentiation.As mentioned, our case of B/T-lineage co-dominant MPAL harbored NF1 and BCOR1 mutations and NUP98::MLLT1.NF1 is commonly mutated in B-ALL and reported in one case of B/T MPAL

Table 1
Clinicopathological features of B/T MPAL.
MRD: minimal residual disease in BM detected by flow cytometry at day 29; N, no; Y, yes; M, male; WBC, white cell count; Hb, hemoglobin; Plt, platelet, PB, peripheral blood; BM, bone marrow; LA, lymphadenopathy; LN, lymph node; CR1, first complete remission.R. Zheng et al.complete remission at the end of induction without minimal residual disease (MRD < 0.01 % by flow cytometry on a BM sample) and was in first complete remission at the last follow-up (36 months after diagnosis) (Table

Table 2
Immunophenotype of B/T MPAL by flow cytometric analysis.